TW201415289A - System and method for providing gesture-based user indentification - Google Patents

Abstract

A mechanism for allowing a user to prove their identity on touch-based devices employing the use of a touch surface in firmware-controlled environments is discussed. The user may prove his or her identity by entering a series of strokes on the touch-based device to form a word or image. Characteristics of the entered strokes such as stroke order and stroke direction are compared to stored stroke characteristics that were gathered from a drawing of the same word or image during a user enrollment process. If the stroke characteristics comparison is acceptable, the user identity is verified.

Description

System and method for providing gesture-based user identification [Related application]

This application is related to and claims the U.S. Patent Application Serial No., entitled "System and Method for Providing Gesture-Based User Identification", filed on September 12, 2012. The benefit of the U.S. Patent Application Serial No. 13/611,992, the disclosure of which is incorporated herein in its entirety.

The present invention relates to systems and methods for providing gesture-based user identification.

Traditionally, the user's identity has been proven by something owned by the user, something that the user knows, or the user's own (ie, the physical characteristics of the user). For example, a user may use a hardware or software license or badge that they have to prove their identity. The user may also prove his identity by something he knows, such as his mother's maiden name or password. Alternatively, the user can use the use The physical characteristics of the person (such as fingerprints, retinal scans or DNA samples) to prove their identity. These different mechanisms for proving the identity of the user can be utilized to control access to today's computing devices. However, many of these techniques present challenges to mobile users due to platform constraints such as the lack of a physical keyboard or necessary interface for input devices. Depending on the desired organization, such constraints make it difficult or impossible for the mobile device user to enter the necessary information to prove identity. For example, in the absence of a suitable interface, the mobile device will not support a fingerprint sensor or smart card, and a user accustomed to a physical keyboard may have difficulty entering a password via a touch screen displayed on the screen. In addition, existing methods specifically designed for use in touch-based devices to prove user identity (such as handwriting recognition processes) often require too much storage and/or computational power in a limited resource environment (such as a user identification process). The case of being performed by the firmware of the computing device) works effectively.

Embodiments of the present invention provide a mechanism for user identification on a touch-based device. More specifically, embodiments of the present invention allow for receiving and analyzing a gesture-based "password" input by a user on a touch surface with a series of gestures or strokes while simultaneously drawing or tracing a word or image. Instead of recording the input word or image as a gesture password as a whole, features related to the constituent gestures constituting the input word or image, such as stroke direction, stroke order, and stroke connection, may be recorded as gesture passwords. The recorded gesture password record has a small size and can be processed using a limited code during the verification process, making the gesture code of the present invention particularly suitable for use as Executed by a portion of the identification process performed by the firmware in the computing device.

In an embodiment, a computing device provides gesture-based user identification. The computing device includes a touch surface that accepts input gestures from a user, and a processor that is configured to execute instructions for user identification. The instructions prompt the user to enter a word or image via the contact surface. The word or image input is a separate stroke, and the features of the strokes together form a stroke password. The stroke cipher is stored as information about the stroke direction and stroke order of each stroke in the individual strokes forming the stroke cipher. The stored stroke password is associated with the user. Following the storage of the stroke code, the instructions further receive words or images entered as separate strokes via the touch surface, the features of the individual strokes collectively forming a stroke verification password. The stroke verification password includes information on the stroke direction and the stroke order of each stroke in the individual strokes in the stroke forming the stroke verification password. The stroke verification password is compared to the stored stroke password. The instructions also identify the user who entered the stroke verification password based on a comparison between the saved stroke password and the stroke verification password.

In another embodiment, a computer implemented method for providing gesture-based user recognition prompts a user to input a word or image via a touch surface that is capable of accepting an input gesture from a user. The word or image input is a separate stroke, and the features of the strokes collectively form a stroke password. The method also stores the stroke password entered by the user. The stroke cipher is stored as information about the stroke direction and stroke order of each stroke in the individual strokes forming the stroke cipher. The stored stroke password is associated with the user. The method is also connected to the touch surface after the password is stored. Individual strokes are received, and the characteristics of the individual strokes collectively form a stroke verification password. The stroke verification password includes information on the stroke direction and the stroke order of each stroke in the individual strokes forming the stroke verification password. The method further compares the stroke verification password to the stored stroke password and identifies the user who entered the stroke verification code based on a comparison between the saved stroke password and the stroke verification password.

800‧‧‧ Computing device

802‧‧‧ processor

804‧‧‧ operating system

806‧‧‧ memory

808‧‧‧ Non-electrical storage

810‧‧‧ touch surface

830‧‧‧ Firmware

832‧‧‧ Registration Module

834‧‧‧ verification module

One or more embodiments of the present invention are described in conjunction with the accompanying drawings, in which the claims In the drawings: Figure 1 depicts an exemplary sequence of steps performed by an embodiment of the present invention, which performs a registration process for a user; Figure 2 depicts an exemplary of the steps performed by an embodiment of the present invention Sexual order, which is used to identify the user; FIG. 3 shows an exemplary stroke direction recorded for a separate stroke completed by the user during the registration process or the identification process; FIG. 4 depicts execution by an embodiment of the present invention An exemplary sequence of steps for converting a trace into a stroke feature; FIG. 5 depicts an exemplary framed trace and sample decode available in an embodiment of the present invention; FIG. 6 depicts an embodiment of the present invention Exemplary enhanced illustration of utilization; FIG. 7A depicts an exemplary stroke password comparison that may be utilized with embodiments of the present invention; FIG. 7B depicts another exemplary stroke password comparison that may be utilized with embodiments of the present invention; and FIG. 8 depicts suitable for practice An exemplary environment of an embodiment of the invention.

Embodiments of the present invention provide a mechanism that allows a user to prove their identity on a touch-based device that utilizes a touch screen or touch surface for use in a firmware control environment. The user can prove his or her identity by entering a series of strokes on the touch-based device to form a word or image. The features of the entered strokes, such as the stroke order and stroke direction, are compared to the stored stroke features collected from the same word or image during the user registration process. If the stroke characteristics match or meet other criteria, the user's identity is verified.

The computing device is initialized by a firmware included in the device, and the firmware provides a series of software services that facilitate the activation of the operating system, and provides a smaller subset of such services, after the operating system has been started Can continue to use. The tough system has been written to software on a read-only memory (ROM) module, including but not limited to ROM, PROM, EPROM, EEPROM, and flash memory (hereinafter collectively referred to as "ROM"). Among the various services, the firmware is particularly responsible for the operation of the computing device until the boot process can be run, which will be used to load the operating system of the computing device. Recalling the body. Once the loading is complete, the operating system is responsible for the normal operation of the computing device, although providing certain services after loading the operating system for security reasons may require transferring control from the operating system back to the firmware. Due to the cost of the ROM module, the amount of space available in the computing device to store the firmware ROM is limited. Thus, the user identification process described herein has the advantage that embodiments of the present invention provide a user identification process that can be practiced by firmware that can be based on touch even in a resource-constrained environment in which the firmware is typically operated. Executed on the device.

Conventional techniques that have been developed to prove identity on touch-based computing devices are not particularly well suited for execution by firmware, which lacks more resources or requires more resources than existing resources when executing firmware. For example, a 9 or 12 sash of a smart phone (which decodes the pattern into a frame number) suffers from security difficulties because the pattern used in practice is relatively small, the user only starts in a few places, and the patterns Use up to four adjacent boxes at most. Handwriting recognition provides better security than the 9 or 12 sash and works well with touch-based interfaces. However, handwriting recognition is more difficult to implement using firmware because it is computationally intensive, that is, the handwriting recognition requires a large amount of code to perform input handwriting processing and requires a large database, both of which are in the context of firmware operation. Cost limit factor. In contrast to these prior art techniques, embodiments of the present invention store gesture-based password records that are small in size and that can be processed using limited code.

The user authorization process of the present invention begins with a registration process during which the user provides and subsequently confirms a gesture based password. Therefore, the registration is used by the user to share the secret with the system (in this case In the case of a stroke/stroke-based password, the system will then use the secret to verify the identity of the user. The registration process may be initiated by the current user who already has the necessary administrative privileges to save the password, or if the target user being registered does not have sufficient privileges to initiate the process, the registration process may need to be authorized by the administrator. The password can be stored to control subsequent access to any of a number of aspects of the computing device. For example, a password can be used to control user access to the computing device's BIOS setup menu.

1 depicts an exemplary sequence of steps performed by an embodiment of the present invention that perform a registration process for a user. The registration sequence begins with the user selecting one or more words or images (step 102). The selected word or image is recommended for the registration process. For example, the registration process of the present invention can provide a user with a word or image to pull down a list or other collection and prompt the user to make a selection. The list can be a word or image from a predefined list, or the registration process can produce a random combination ("red black" or "big job" or "real abs"). The method of registration of such a suggested word has the disadvantage that the word is usually less readable (because it is not chosen by the user), but it has the advantage that no typing is required. Alternatively, in one aspect of the invention, the user may provide the word or the like for registration by entering one or more words or images of their own on a screen keyboard or other keyboard/input device. process. In the case of automatically generating a word/image list, the total number of strokes of these words should preferably be small. In one embodiment, only one or two characters are required for a set of ideogram characters such as those used in Chinese, Korean, and Japanese.

Once one or more words or images have been selected, the registration process is The user is displayed and prompted to trace it on the touch-based surface of the computing device (step 104). It should be appreciated that the touch-based surface of the computing device is often also the display screen of the computing device, but in alternative embodiments, the touch surface can be a touch-sensitive input-independent surface that is different from the display screen/surface. Thus, where the terms "touch screen" or "touch surface" are used in this description, it should be understood to include also a touch-based surface that is different from the main display screen/surface of the computing device. During the registration process, the word or image may be enlarged to make the tracing process easier. The tracing can be performed by finger strokes completed on the touch surface, and the movement of the finger when the screen is touched is periodically sampled and recorded. For example, in one embodiment, coordinates from movement can be sampled and coordinates relative to the upper left edge of the screen recorded. It should be understood that the above process can also be performed without magnifying words or images. In an alternate embodiment, the selected word or image can be traced according to the user's memory without displaying the word or image on a touch-based screen.

The registration process uses words or images as visual representations of the user. The use of words or images creates a series of strokes in a simplified manner, and the user will enter the strokes as a gesture/stroke password during both the registration process and the subsequent verification/recognition process. For example, if the DOG is the word, the user remembers the DOG (relatively simple mental image), but draws the bottom-up-right-down-left, right-down-left-up, left-down-right-up (short) or the like instead of typing DOG. As described in further detail below, the registration process does not trace the shape of the individual characters, but rather captures the direction in which the fingers are traveling and other features. You can also use the quadrant and path simplification to explain the "wide finger" question. The question is also explained further below.

It will be appreciated that the depiction of words or images during registration can be performed in many different ways known in the art, such as tracing performed on a touch surface using finger strokes, stylus strokes, or other touch-based gestures. In alternative embodiments using a touch-free screen, other methods such as mouse tracing may be utilized. The feature of the individual stroke of the user tracing the selected word or image is captured by the registration process and converted to a stroke cipher in a manner described further below (step 106). Subsequently, the drawing area of the touch surface is cleared, and the user can be prompted to repeat the process and confirm the stroke of the stroke password without displaying the pre-rendered image (step 108). Optionally, in an alternate embodiment, the application also provides stroke guidance (such as a stroke indicator) such as a screen arrow and user feedback, without the user confirming the original traced drawing based entirely on memory. The user's tracing stroke features for the word or image are compared to the stroke characteristics of the user's subsequent confirmation drawing of the word or image (step 109). If there is no match between the two attempts (step 110), the user may be prompted to try again or may terminate the process depending on how the registration process is performed. On the other hand, if the stroke features match, the stroke password is confirmed (step 112), associated with the user, and saved to the non-electrical memory for later retrieval (step 114).

It should be noted that the registration process of the present invention stores the features of the strokes that collectively form a word or image as a stroke code in a non-electrical memory rather than storing the complete image that is drawn. The relationship between the strokes used and the words is not recorded. The stored features may include a stroke order, a stroke direction. The features may also include a stroke connection (ie, the finger between the strokes is No mention). Additional stroke features can also be saved, such as stroke timing (how long the user spends completing the stroke) and stroke strength (pressure or thickness of the stroke). In the practice of the present invention, the user identity is then evidenced by the stroke characteristics of the strokes that the user performed when drawing the word or image, rather than by comparing the input to the stored complete image.

After registration, the saved stroke password from the registration process is then used to identify the user. This identification may involve asking the user to select who they will log in with, and then having the user who is logged in enter the verification password, comparing the characteristics of the verification password with the characteristics of the saved stroke password, and the saved stroke passwords. Features are associated with the claimed identity. Alternatively, the identification process may involve causing the user to enter a verification password that is compared to the characteristics of all of the saved passwords.

2 depicts an exemplary sequence of steps performed by an embodiment of the present invention for identifying a user. The identification process can begin with the user drawing a word or image (step 202). The features of the individual strokes created by the user for drawing words or images are captured and collectively form a verification password (step 204). Subsequently, the verification password is compared to all of the saved stroke ciphers in an attempt to identify an exact match or satisfaction with the defined criteria, the satisfaction of the defined criteria is somewhat more flexible and does not require an exact match (step 206). If the comparison result satisfies the condition (step 207), the user is identified as the user associated with the saved stroke password (step 208). Alternatively, if the result of the comparison does not satisfy the condition (step 207), the user is unrecognized and the user may be denied access to all or a portion of the computing device controlled by the password (step 210). In an alternate embodiment, rather than comparing the verification password to all of the saved passwords, the user can provide the identity they claim, and the identification process can only compare the verification password to the saved stroke password, the stroke password. Associated with the individual claimed by the user.

As described above, the captured stroke features used to collectively form the stroke and verify the password may include stroke direction, stroke order, and in some cases include stroke connections or other stroke features. For example, Figure 3 illustrates an exemplary stroke direction that can be stored for each stroke that a user completes when drawing a word or image during the registration process. The stroke direction can be recorded as upper (302), lower (304), left (306), and right (308). Similarly, the stroke direction of a single stroke can be recorded as the lower left (310), the lower right (312), the upper left (314), and the upper right (316). It will be appreciated that other orientations besides or in place of those depicted in FIG. 3 may be stored without departing from the scope of the invention. When the full gesture fails to cross the minimum amount of distance on the touch surface, the stroke can also be recorded as a point (318). It also captures the order in which strokes are entered and the connections between strokes. The gesture record performed by embodiments of the present invention is not sensitive to relative character or screen positioning. That is, whether the drawn characters overlap or are correctly positioned or their position on the screen is not important. In addition, the captured gestures can be drawn at an angle. The captured stroke information can be stored in a relatively compact form, for example, in the form of 5 strokes per stroke and 2-4 strokes per character. Although this roughly causes the storage of each character to be twice the storage of the text password characters, it is much smaller than the storage capacity of the complete gesture capture and the gesture processing code is small, which is equivalent to executing the password. The amount of code needed to hash.

The registration process and the identification process described above with respect to FIGS. 1 and 2 process the "strokes" entered by the user by capturing information collected from the user's registration of one or more words or images and Subsequent confirmation drawing, the word or image is converted into a series of strokes or defined movements. These strokes can then be compared to the recorded stroke set. The stroke can be a finger event or a gesture event. A finger event occurs when the user touches the touch surface (finger down) or stops touching the touch surface (finger up). For the purposes of this description, the terms "touch surface and touch screen" refer to surfaces that detect the movement of a finger. Although this description uses words such as "hand gesture", "finger down" and "finger up", it should be understood that similar actions can be made using mouse movement, mouse button down, mouse button up or other types of pointing devices. behavior. In an embodiment of the invention, a stroke or gesture event occurs after the finger down event but before the finger up event, as the user moves the finger across the touch surface. As discussed above, strokes can be categorized into nine basic types: point, move to the left, move to the right, move up, move down, move to the top, move to the left, move to the bottom, and move to the left. However, it should be appreciated that additional gestures may be added or used to replace some or all of these gesture types. For example, short or long strokes or different stroke angles can be extrapolated to distinguish similar gestures.

In an embodiment of the invention, when the user makes a gesture, the position of the finger can be detected relative to the position on the touch surface (such as the upper left) as the user's finger moves over the touch screen/surface. When the finger is down, the recorded samples of the X and Y screen/surface coordinates of the finger are captured and collectively referred to as a trace. The trace described below with reference to Figure 4 converts the trace into a series of one or more strokes. A feature set of all strokes from all traces captured by the user when tracing/rendering one or more words or images creates a stroke or verification password, and the capture process is the same for both types of gesture passwords.

4 depicts an exemplary sequence of steps for converting a trace into a stroke that, in accordance with an embodiment of the present invention, collectively form a stroke or verification password. The tracing process (400) begins by calculating a "boundary quadrilateral" (402) for each tracing. The boundary quad is created in three steps. As a first step, the smallest rectangle containing all the trace coordinates is calculated. By definition, each side of this rectangle will intersect at least one trace coordinate. In the second step, for each corner of the rectangle, the closest trace coordinates along either of the two rectangular sides attached to the corner are found. In the third step, the boundary quadrilateral is constructed based on the four coordinates found in the previous step. The bounding rectangle is the smallest rectangle that can enclose the entire de-skewed trace (de-skew is discussed further below). Subsequently, the trace is checked by checking the "X size" and "Y size" of the trace to verify that it is not a "dot" stroke (403). The X size is the difference between the largest X coordinate in the trace and the smallest X coordinate in the trace. The Y size is the difference between the largest Y coordinate in the trace and the smallest Y coordinate in the trace. If neither the X size nor the Y size exceeds the predetermined "point threshold", the stroke is composed of dot strokes. Such point analysis should be a condition in which the user puts his finger down and then lifts his finger without moving his or her finger a significant distance, such as the condition of a point above the period or lowercase "i". Although the user desires a point, the trace may contain several sets of coordinates. For example, when using a mouse device, a small movement can produce a small change in coordinates. Or, when using the touch screen, the finger actually touches several points on the touch screen. The point threshold helps distinguish between this condition and the situation when the user has made a gesture. In addition, since the user has lifted his finger, the stroke of "stroke completion" is recorded to distinguish the point stroke from any subsequent stroke. If it is a point stroke, the processing of the trace is skipped and all remaining traces continue to be processed (414).

If the trace is not a point, it is de-skewed (404). For example, a user can draw a stroke at an angle. There are various well-known transforms that can be used to de-skew a set of coordinates. One such method compares the angle produced by the bottom and left edges of the boundary quadrilateral with the angle produced by the bottom and right edges of the boundary quad. If they are roughly the same, the trace can be skewed and the coordinates of the trace can be adjusted such that the left and right edges of the boundary quadrilateral are at 90 degrees (vertical).

After the trace has been de-skewed, the bounding rectangle is calculated (406). As mentioned above, the "boundary rectangle" is the smallest rectangle that can enclose the entire de-skew trace. The "normalized boundary rectangle" is composed of a boundary rectangle whose leftmost X coordinate is 0 and the topmost Y coordinate is 0.

After the calculation of the bounding rectangle, the trace is converted to "framed trace". The framed traces are formed by dividing the total area of the bounding rectangle into 3 x 3 grids and converting the traces into sashes (408), each of which covers approximately 1/3 of the total width and height. Then, by where the recorded trace coordinates fall into A grid square is used to process the trace. The resulting record is called a "framed trace." For convenience, the frames may be numbered sequentially from top to bottom, from left to right, starting with 1. Thus, a 3x3 grid can have blocks 1 through 9. It will be appreciated that the size of the grid can be varied as needed to give a larger definition without departing from the scope of the invention. It will be appreciated that in alternative embodiments, the framed traces may be formed using "n x n" or "n x m" or "m x n" grids, where "n" is not equal to three.

Once the trace has been framed, it is simplified (410). Due to the mouse shake or "fat finger", the framed trace can be turned to the adjacent box before reaching the desired destination. These biases are well understood and predictable. By searching for a particular pattern in a framed trace and replacing it with a simplified pattern, the process of creating a stroke from the trace is easier.

For example, Figure 5 depicts an exemplary framed trace and sample decoding that may be utilized with embodiments of the present invention. In Figure 5 , the grid (500) is superimposed on the uppercase "M" trace (502) drawn by the user. The grid is a 3 x 3 grid with 3 columns (504), (506), (508) and individual unit cells labeled "1 to 9 ("quadrant"). In one example, the letter "M" can be drawn as a single trace, in quadrant order (520) "7-4-1-2-4-5-8-5-6-2-3-6-9" Cross the quadrant of the grid 500. The tracing conversion process of the present invention interprets strokes/characters of the order "7-4" and "7-4-1" into upward strokes. Trace the new strokes from the "up" to "bottom right" directions. Strokes/character "1-4-5-8", "1-2-4-5-8", "1-2-5-4-8" and "1-5-4-8" (depending on use) The degree of straightness of the gesture and the quadrant of the gesture can be interpreted as the "bottom right" stroke by the trace conversion process of the present invention. Similarly, the change to the "upper right" direction will also start a new stroke, and the quadrant order "8-5-3", "8-5-6-3", "8-6-5-3" can all be used. The tracing process of the invention is interpreted as a "top right" stroke. Similarly, the change to the "down" direction and the quadrant order "3-6" and "3-6-9" will be interpreted as downward strokes.

In FIG. 5 , an exemplary coding portion (525) depicts an exemplary stroke direction interpretation that may be utilized by the trace conversion process of the present invention. As depicted, the four identified strokes in the trace (502) can be reduced to the start and end quadrants and the associated order and direction information saved as part of the stroke password during registration (530). For the exemplary trace (502), this information may therefore include "7-1", that is, up, then "1-8", that is, the lower right, then "8-3", that is, the upper right, then "3- 9", that is. The process is the same during the identification process, except that the captured information is compared to the saved information to verify the identity of the user. It will be appreciated that the specific interpretation of the tracing information used to illustrate inaccurate user plots will vary from implementation to implementation, depending on how much directional deviation is considered to be a stroke or a direction.

Some users are more proficient in making the gestures described herein than others. Thus, in one embodiment of the invention, an enhanced gesture can be used to assist the user in tracing and drawing a word or image during the registration process. For example, as depicted in FIG. 6, an alternative to assist the user in order to strengthen the displayed character or image "schematic" (six hundred) (display of the letter "M") is "a schematic" aspect, the use of such The stroke drawn by the person is detected and drawn as a trace of the schematic word or image. For example, a pre-drawn raster is superimposed on an enlarged character character depending on the character value. The grid size is calculated based on the size of the character characters plus the extra bounding area on all sides. For example, a narrow character can occupy only a single line, while a wide character can occupy all 3x3 rasters, and the like. Other grid mating methods are also possible within the scope of the present invention. When using the enhanced gesture, as the user traces the letters, stylized arrows are superimposed on the trace to show the detected strokes (602), (604), and (606), using the same process as described above. The exception is a grid that uses pre-computed rasters instead of trace-based bounding rectangles in real time. In another alternative embodiment, an empty grid may exist when the user enters and verifies the stroke password, and even when the stroke password is entered during the identification. This program enhances the user's knowledge of the strokes that make up the stroke code and improves the reproducibility of their strokes when drawn.

Although the above description has focused on verifying a strict comparison between the signature characteristics of the password and the characteristics of the saved stroke password, it should be appreciated that embodiments of the present invention may also be implemented to allow for greater flexibility in determining whether a match has occurred. Sex. For example, in one embodiment, the stroke order or direction can be interpreted in a non-rigid manner using a more flexible standard. Thus, users who draw strokes in a different order or in the opposite direction may still be considered a match. Although this reduces the security of the password (because more possible matches are allowed), it reduces the confusion caused by the fact that some characters can be drawn in a variety of ways, even by the same user. FIG. 7A depicts an exemplary stroke password comparison in which strokes in a single character can be entered in any order as long as other features match. Therefore, the original stroke order 1-2-3-4 drawn for the letter "M" (702) during registration will be considered as a match for the verification stroke order 4-1-2-3 (704) because the other stroke features are the same. . It will be appreciated that the same result can be achieved if the 4-1-2-3 (704) order is drawn during the registration period and the 1-2-3-4 (702) order is drawn during the verification period. Similarly, FIG. 7B depicts an exemplary stroke password comparison in which strokes that are exactly opposite each other are considered identical. That is: upper = lower, left = right, lower right = upper left and lower left = upper right. For example, in the case where an exact match is required, the letters (712) and letters (714) will not be considered a match due to the difference in direction. However, using the reversed alternatives depicted, the drawn letters will be considered to meet the comparison criteria because, although the directions of the strokes are different, they are mirror strokes. It will be appreciated that the embodiments described in Figures 7A and 7B can also be combined when determining the match between the saved stroke password and the verification password, and other types of flexible "matching" can be implemented.

FIG. 8 depicts an exemplary environment suitable for practicing embodiments of the present invention. The exemplary computing device 800 is equipped with a touch surface 810 through which a user inputs acceptable gestures and strokes, such as finger movements. The computing device 800 can be a tablet computing device, a cell phone, a smart phone, a PDA, a laptop, a portable Internet device, a desktop computer, a mobile computing device, or other computing device, equipped with a touch surface or Touch surface communication and the ability to perform the functions described herein to register and verify a gesture based password. Computing device 800 includes a processor 802 for executing instructions and an operating system 804. The computing device further includes a memory 806, such as a random access memory (RAM), and includes or has access to a non-electrical memory such as a hard disk drive. Computing device 800 further includes a firmware 830 that is responsible for device initialization and initially loading OS 804 into memory. The firmware 830 includes a registration module 832, which is responsible for performing the processing described in FIG. 1 herein . The embodiment of the present invention inputs and saves a stroke password by the registration module. The firmware 830 further includes a verification module 834 for performing the processing described in FIG. 2 , and the verification password is accepted by the verification module and compared with the saved stroke password to attempt to perform user identification. It will be appreciated that the registration module 832 and the verification module 834 can be combined into one module or divided into additional modules without departing from the scope of the invention.

Although the description incorporated herein has focused on touch-based devices in which the firmware performs the recognition process, it should be understood that embodiments of the invention are not so limited. For example, embodiments of the present invention may also be implemented to accept mouse drawing/draw movement instead of finger movement on a touch surface. Alternatively, instead of practicing the invention by the firmware of the computing device, the identification techniques described herein can be performed by an application or process under the control of the operating system and can verify the identity of the user, thereby controlling the user to various applications. General access to information, equipment or devices (ie, as required for general login).

Some or all of the embodiments of the present invention may be provided as one or more computer readable programs or code embodied on or in one or more non-electrical media. Such media may be, but are not limited to, hard drives, optical disks, digital versatile discs, ROM, PROM, EPROM, EEPROM, flash memory, RAM or magnetic tape. In general, computer readable programs or code can be implemented in any computing language.

Since some changes can be made without departing from the scope of the invention, it is desirable to incorporate or incorporate in the above description. Everything shown is interpreted as illustrative rather than literally. A person skilled in the art will recognize that the order of the steps and the architecture depicted in the figures can be varied without departing from the scope of the invention, and the examples incorporated herein are specific examples of numerous possible descriptions of the invention.

The above description of the exemplary embodiments of the invention is intended to be illustrative and not restrictive Modifications and variations are possible or can be derived from the practice of the invention. For example, although a series of acts have been described, the order of the acts may be modified in other embodiments consistent with the principles of the invention. Furthermore, non-dependent actions can be performed in parallel.

Claims (27)

A non-electrical medium retaining computer executable instructions for providing gesture-based user identification, the instructions causing, when executed, causing at least one computing device to prompt a user to input a word or image via a touch surface The touch surface is capable of accepting an input gesture from the user, the word or image input being a plurality of individual strokes, the features of the plurality of individual strokes collectively forming a stroke code; storing the stroke password input by the user, The stroke password is stored as information about the stroke direction and the stroke order of each of the individual strokes in the plurality of individual strokes forming the stroke code, and the stored stroke password is associated with the user; After storing, receiving, by the touch surface, a word or image as a plurality of individual strokes, the plurality of individual strokes collectively forming a stroke verification password, the stroke verification password including in the plurality of strokes forming the stroke verification password Information on the stroke direction and stroke order of each stroke in the individual strokes; Verify the password with the password stored stroke comparing; stroke and based on the stored password to identify the user to enter the verification code stroke with the stroke of the comparison between the password authentication. For example, in the medium of claim 1, the firmware is executed. For example, the media in the first paragraph of the patent application, wherein the instructions are The executing further causes the at least one computing device to prompt the user to re-enter the stroke password; and to verify the match by comparing the stroke password with the re-entered password before storing the stroke password. The medium of claim 3, wherein the user initially inputs the stroke code by tracing one of the words or images, and draws the word or image by not displaying the word or image. To re-enter the stroke password. The medium of claim 1, wherein the pre-drawn grid is superimposed on a magnifying character character when the user is inputting at least one of the individual strokes that collectively form the stroke cipher. As for the medium of claim 4, as a user traces the character, a stylized arrow is superimposed on the trace. The medium of claim 1, wherein when the user is inputting at least one of the individual strokes that collectively form the stroke code, a pre-drawn grid is superimposed on the touch surface and the touch The surface does not display the representation of the stroke password. The medium of claim 1, wherein each individual stroke is initially recorded as a series of touch surface coordinates for capturing a one-finger event or a potential event, which occurs when a user touches the touch surface or stops touching the touch surface. A one-finger event occurs when a user moves their finger across the touch surface between finger events. The medium of claim 8, wherein one of the traces for a potential event consists of a set of x and y touch surface coordinates, and the The trace is processed into a series of one or more strokes. For example, the medium of claim 9 of the patent application, wherein the process will draw one of the traces at an angle to deviate. The medium of claim 9, wherein the processing: calculating a boundary rectangle for the trace, the boundary rectangle being a minimum rectangle of all the coordinates that can surround the trace; a total area of the boundary rectangle Divided into n x n grids; record which grid squares are occupied by the trace; and adjust the record of the occupied grid square based on predetermined criteria to produce a simplified frame number pair indicating a stroke direction. For example, in the media of claim 1, the stroke password and the verification password also include at least one of stroke connection information, stroke timing information, and stroke strength information. A computing device for providing gesture-based user identification, the computing device comprising: a touch surface that accepts input gestures from a user; and a processor configured to execute instructions to: prompt a user via Inputting a word or image on the touch surface, the word or image input is a plurality of individual strokes, and the features of the plurality of individual strokes together form a stroke code; storing the stroke password, the stroke password is stored as being related to forming the stroke password Information about the stroke direction and the stroke order of each stroke of the individual strokes in the plurality of individual strokes, and the stored stroke password is associated with the user; After the storing, receiving, by the touch surface, a word or image as a plurality of individual strokes, the plurality of individual strokes collectively forming a stroke verification password, the stroke verification password including the plurality of strokes regarding the formation of the stroke verification password Information about the stroke direction and stroke order of each stroke in the individual strokes; comparing the stroke verification password with the stored stroke password; and based on the saved stroke password and the stroke verification password The comparison identifies the user who entered the stroke verification code. The computing device of claim 13, wherein the execution of the instructions is a firmware. The computing device of claim 13, wherein the stroke password and the verification password further comprise at least one of stroke connection information, stroke timing information, and stroke strength information. A computer-implemented method for providing a gesture-based user identification for a computing device, the computer implementation method comprising: prompting a user to input a word or image via a touch surface, the touch surface being capable of accepting from the user Entering a gesture, the word or image input is a plurality of individual strokes, and the characteristics of the plurality of individual strokes together form a stroke code; storing the stroke password input by the user, the stroke password is stored as being related to forming the stroke password Information about the stroke direction and the stroke order of each stroke of the individual strokes in the plurality of individual strokes, and the stored stroke password is associated with the user; After the storing, receiving, by the touch surface, a word or image as a plurality of individual strokes, the plurality of individual strokes collectively forming a stroke verification password, the stroke verification password including the plurality of strokes regarding the formation of the stroke verification password Information about the stroke direction and stroke order of each stroke in the individual strokes; comparing the stroke verification password with the stored stroke password; and based on the stored stroke password and the stroke verification password The comparison identifies the user who entered the stroke verification code. The method of claim 16, wherein the execution of the instructions is a firmware. The method of claim 16, further comprising: prompting the user to re-enter the stroke password; and verifying the match by comparing the stroke password with the re-entered password before storing the stroke password. The method of claim 18, wherein the user initially inputs the stroke code by tracing the display of the word or image, and draws the word or image by displaying the word or image without displaying the word or image. To re-enter the stroke password. The method of claim 16, wherein a pre-rendered grid is superimposed on a magnifying character character when the user is entering at least one of the individual strokes that collectively form the stroke cipher. For example, the method of applying for the scope of patent 20, along with a user The character character is traced and a stylized arrow is superimposed on the trace. The method of claim 16, wherein when the user is inputting at least one of the individual strokes that collectively form the stroke code, a pre-drawn grid is superimposed on the touch surface and the touch The surface does not display the representation of the stroke password. The method of claim 16, wherein each individual stroke is initially recorded as a series of touch surface coordinates capturing a one-finger event or a potential event, which occurs when a user touches the touch surface or stops touching the touch surface. A one-finger event occurs when a user moves their finger across the touch surface between finger events. A method of claim 23, wherein one of the traces for a potential event consists of a set of x and y touch surface coordinates, and the trace is processed into a series of one or more strokes. The method of claim 24, wherein the process is to trace one of the traces at an angle. The method of claim 24, wherein the processing: calculating a boundary rectangle for the trace, the boundary rectangle being a minimum rectangle of all of the coordinates surrounding the trace; a total area of the boundary rectangle Divided into n x n grids; record which grid squares are occupied by the trace; and adjust the record of the occupied grid square based on predetermined criteria to produce a simplified frame number pair indicating a stroke direction. For example, in the method of claim 16, wherein the stroke password and the verification password also include stroke connection information, stroke timing information, and pen At least one of the intensity information.